The Swedish Macroeconomic Policy Framework Lars Calmfors Institute for International Economic Studies, Stockholm University

Transcription

1 1 The Swedish Macroeconomic Policy Framework Lars Calmfors Institute for International Economic Studies, Stockholm University Chapter in Handbook in Swedish Politics (ed. Jon Pierre),Oxford University Press, preliminary version 30 July, 2013.

2 2 Sweden s macroeconomic policy framework has recently come to been seen as a role model for other countries. The decade before the international economic crisis erupted in 2007/08 was characterised by low and stable inflation (Figure 1) as well as small fluctuations in economic activity. There were fiscal surpluses (Figure 2) and government debt (Figure 3) was reduced substantially to levels far below the Maastricht benchmark of 60 per cent of GDP. Sweden has also weathered the international crisis well. Public finances have remained in good shape. The rise in unemployment (Figure 4) has been limited. The recent Swedish performance stands in stark contrast to earlier developments. In the 1970s and 1980s, Sweden was trapped in a devaluation cycle, which culminated in a deep financial crisis when a real estate price bubble burst in the early 1990s. The crisis involved large fiscal deficits (Figure 2) and huge rises in unemployment (Figure 4). In the aftermath of the 1990s crisis, a new macroeconomic policy framework was established. In the monetary policy area, it involved the adoption of inflation targeting and the granting of more independence to the central bank. In the fiscal policy area, it meant a stricter framework, imposing more discipline, and pension reform. Section 1 discusses the monetary policy framework and Section 2 the fiscal policy framework. Both sections consist of three parts: (a) a review of the current framework; (b) an analysis of how the framework was established; and (c) a discussion of future challenges. Section 3 draws some general conclusions on how crisis experiences can trigger reforms and the roles played by academic research and international influences. 1 The monetary policy framework 1.1 The current framework According to the Constitution, the central bank (the Riksbank) is in charge of monetary policy. The bank s activities are regulated in the Riksbank Act. It stipulates that the Riksbank s objective is to maintain price stability, but does not define this concept explicitly. The bank itself has adopted an inflation target of an annual two per cent increase in the CPI. The Act does not specify any output or employment objectives. However, the Government Bill proposing the currently existing monetary policy framework pointed out that the Riksbank, as a member of the European Central Bank System (ECBS) is, without prejudice to the price stability objective, obliged to support the EU s general economic policies which seek among other things to achieve balanced economic growth and full em-

3 3 Figure 1 CPI inflation Note: The broken line indicates the 2 per cent inflation target. Source: OECD.

4 4 Figure 2 General government net lending, per cent of GDP Note: General government net lending is the difference between the sector s revenues and expenditure as defined in the national accounts. The broken lines indicate the surplus target of 1 per cent of GDP and the EU deficit ceiling of 3 per cent of GDP respectively. Source: SCB, Sweden.

5 5 Figure 3 Government debt, per cent of GDP Note: Consolidated general government gross debt is general government total debt after all internal claims and liabilities in the sector have been netted out. General government net debt is the sector s gross financial debt minus its financial assets (including claims on the private sector). The horizontal broken line indicates the EU debt ceiling of 60 per cent of GDP. Sources: Consolidated gross government debt: Ameco. General government net debt: SCB, Sweden.

6 6 Figure 4 Unemployment, per cent of the labour force Source: OECD.

7 7 ployment (Proposition 1997/98:40). The minutes from the meetings of the bank s Executive Board (see below) also show that the bank sees it as an objective to stabilise output and employment around their equilibrium (sustainable) levels. This policy is often referred to as flexible inflation targeting. According to established theory, such a policy does not involve any conflict in the long term between the price stability and output (employment) targets, as the equilibrium output and employment levels are defined as the ones consistent with stable inflation. The Riksbank has a high degree of independence from the political system. First, the bank has institutional independence. The government is prohibited from giving instructions to the bank, and the bank is not allowed to take instructions from the government. 1 Second, the Riksbank s Executive Board, which takes all policy decisions, have personal independence from the political system. The six members are appointed by the bank s General Council), which is elected by the Parliament. The Council s members are affiliated to political parties, but only some are MPs (currently both the Chair and the Vice Chair are not) and some are politicians who are no longer active. The idea is to have the appointment procedure to the Executive Board at arm s length from day-to-day politics. Board members are appointed for five or six years. They cannot be MPs, ministers or employees in the government offices or at the national level of the political parties. They should be persons with a broad experience of society and economics. The General Council can retire a member of the Executive Board if the member no longer fulfils the conditions required for the performance of his duties or if he has been guilty of serious misconduct, but the decision can be appealed. 2 Third, the Riksbank has considerable economic autonomy. The Parliament decides on a proposal from the bank s General Council how much of the annual profit that should be paid to the government budget, but the Riksbank Act sets minimum levels for equity and regulates how deviations from them should be corrected. 1 As it is the government that decides on the exchange rate system, it could in principle peg the Swedish krona to the currency of another country (or set of countries) with high inflation, which could make it impossible for the central bank to achive its inflation target. This may seem as an inconsistency. It has, however, been argued that since the Riksbank is responsible for the implementation of the exchange rate regime, it can always adjust the exchange rate parity in a way that is consistent with its price stability target (Proposition 1997/98:40). 2 The Governor can appeal to the European Court of Justice and the other Board members to the Swedish Supreme Court.

8 8 It is a challenge how to hold an independent central bank accountable for its policy without interfering with its independence. The Riksbank Act does not require the Executive Board to publish minutes from its meetings, but the Board itself has chosen to do so. The law, however, requires the bank to deliver an account of its policy to the Parliament twice a year. There are also regularly open hearings with the Governor and other members of the Executive Board in the Parliament s Committee on Finance. It publishes an annual evaluation of the bank s performance. The committee also commissions evaluations of the bank s performance by leading foreign monetary policy experts. 1.2 The establishment of the monetary framework The current monetary policy framework is very different from the one in the 1970s and 1980s. Sweden then tried to maintain a fixed exchange rate, first vis-á-vis the D-mark and later vis-ávis a trade-weighted basket of currencies. The exchange rate peg was supposed to anchor domestic inflation to that abroad. This did not happen, however, as the fixed exchange rate policy had low credibility. High inflation expectations repeatedly triggered high wage increases, eroding Sweden s international competitiveness and causing unemployment in mainly the tradables (manufacturing) sector. The wage increases were from time to time (1976, 1977, 1981 and 1982) accommodated through currency devaluations, which fed inflation expectations further, thus triggering new rounds of high wage increases and keeping the inflation-devaluation cycle alive. 3 An important role in the economic policy debate in the 1980s was played by the Economic Policy Group at the Center for Business and Policy Studies (SNS Konjunkturråd), a group of independent academic economists. The Group argued that the government should declare the fixed exchange rate as the overriding objective of macroeconomic policy with full employment being only a subordinate goal. 4 The intention was to make it clear to the parties in the labour market that high wage increases would cause unemployment and falls in profits that would not be accommodated through new devaluations. An explicit commitment to such an exchange rate norm would impose a high reputation cost for the government of 3 See, for example, Jonung (1999) and Finans- och penningpolitiskt bokslut för 1990-talet (2001). 4 The thinking of the SNS group was developed in SNS (1985, 1986, 1987). Similar thoughts had earlier been advanced by e.g. Calmfors et al. (1976), Myhrman (1977), Jonung (1978), Lindbeck (1978) and Calmfors (1979). See also Calmfors (1996) for a survey of the debate.

9 9 abandoning it. This would make the commitment credible and induce the labour market organisations to negotiate wage increases compatible with the fixed exchange rate. Hence the threat of high unemployment would never have to be realised. In the 1980s, the Riksbank was still very dependent on the political system. Policy decisions were taken by the bank s General Council comprised of politicians. Monetary policy objectives were not explicitly defined. The strong incentives for politicians to maintain low unemployment under all circumstances therefore undermined the credibility of the exchange rate norm. For this reason, the SNS Economic Policy Group advocated more independence for the Riksbank. The Group was strongly influenced by the then recent research on the time inconsistency problem of monetary policy, which emphasised how polticians interested in both low unemployment and low inflation had ex post incentives to renege on ex ante announcements of low inflation policy in order to improve employment outcomes. Such policies would, however, in the long term only result in high inflation without any long-run impact on unemployment, as the public would learn about the behaviour of policy makers and thus anticipate their behaviour. 5 Then, in the vocabulary of economists, the so-called Phillips curve, i.e. the relationship between inflation and unemployment, would be vertical. The SNS Group s recommendations first met with strong political resistance. However, during the second half of the 1980s it became increasingly evident that fundamental changes were needed to bring down inflation. The economy was strongly overheated, the boom being fuelled by fast credit growth in the wake of credit market deregulations. In this situation, the ideas of the SNS Group gained increased acceptance. After a failed government attempt in 1990 to legislate against high wage and price increases, the Social Democratic government in 1991 announced that the main economic-policy objective over the coming years must be to permanently reduce inflation and that this task is superior to other ambitions and requirements. 6 The exchange rate peg was then also shifted from a trade-weighted currency basket to the ecu (a weighted average of EU currencies). The SNS Group had an even greater influence on the thinking of the liberal-conservative government which came to power in the 5 The seminal contribution was Kydland and Prescott (1977). See also Barro and Gordon (1983a,b), 6 Proposition 1990:91:100, bil.1, p. 4.

10 10 autumn of In the spring of 1992, it declared that the fixed exchange rate is a crucial and definitive norm for economic policy. 7 The attempt to establish an exchange rate norm was not backed by any major institutional changes. When the norm was introduced, the inflation that it was designed to prevent had already occurred and eroded Sweden s competitiveness. The boom came to an abrupt end in 1990/91 when the property price bubble burst and Sweden entered a deep recession. This triggered expectations of a new devaluation, so the Riksbank had to defend the krona through big interest rate hikes which deepened the downturn. This created an impossible situation and in November 1992 the fixed exchange rate was abandoned. This was amidst a period of general European exchange rate turbulence. Pegging the exchange rate again did not seem a credible option in this situation. At the same time, there was widespread agreement that an anchor was needed to keep inflation under control. There was a strong case for trying to capitalise on the investment in reputation for inflation fighting that had been made during the defence of the fixed exchange rate. Therefore, in January 1993 the Riksbank decided to adopt an annual inflation target of two per cent. The decision was inspired by the monetary policy regimes that had earlier been introduced in Canada and New Zealand. The Riksbank s formal status was analysed by a Government Commission, initially appointed by the Social Democratic government in 1990, but which received new terms of reference and a new composition (of MPs) in 1991 by the liberal-conservative government. The Commission proposed more independence for the bank in its final report in early 1993 (Riksbanksutredningen 1993). The importance of more autonomy for the Riksbank was also underlined by the Lindbeck Commission, another Government Commission, appointed by the liberal-conservative government, consisting of academics and given the remit to propose both policies and institutional reforms to take Sweden out of the economic crisis (Lindbeck Commission 1994). 8 The proposals on a more independent Riksbank were not adopted when they were made. The idea was opposed by the Social Democrats, who came to power again in The main 7 Proposition 1991/92:150, bil I.1, p. 1 8 The Commission was named after its chair, professor Assar Lindbeck. The Swedish version of the report was published in 1993.

11 11 reasons were a worry that monetary policy would be decided by experts, not sharing the political objectives of the government, and a fear of insufficient co-ordination with fiscal policy (Bergkvist and Gradin 1993). However, in 1997 the Social Democratic government changed its position and initiated a fiveparty agreement on more independence for the Riksbank, which in some respects went even further than the proposals by the earlier Government Commission. There are two main reason for why this happened. The first has to do with the Swedish entry into the EU in 1995 (which interestingly cannot just be seen as an exogenous political event changing preconditions for both economic institutions and policies, since the then Social Democratic government s decision to apply for EU membership in 1990 was partly motivated by expectations that this would contribute to macroeconomic stability; Carlsson 2003). As a member of the EU, Sweden was obliged to strengthen the independence of its central bank. The second reason is related to the Swedish decision in 1997 to stay outside the monetary union. This course of action was recommended in 1996 by the Calmfors Commission, another Government Commission, consisting of academics, which had been given the remit to analyse the pros and cons of Swedish membership in the monetary union. 9 The Commission regarded central bank reform, strengthening the Riksbank s independence, as a necessary prerequisite for stable macroeconomic developments if Sweden does not participate in the monetary union. 10 To sum up, the current monetary policy framework in Sweden is the outcome of a long process. Its establishment was a response to an earlier malfunctioning system, which led to a number of academic proposals on reforms, to a large extent inspired by international resarch developments in the monetary policy area. The ideas gained increased acceptance in Sweden around 1990 when the failure of the earlier monetary regime became even more apparent than before. The final transition to the current regime was triggered by two developments: Swedish EU membership, which imposed requirements to do central bank reform, and the Swedish 9 The Commission was named after its chair, professor Lars Calmfors, the author of this chapter. The report was published in Swedish in A translated English version was published in 1997 (Calmfors Commisssion 1997). 10 According to the Government Bill on a new Riksbank Act, the government shared the view of the five-party working group on the status of the Riksbank that it is of particular importance with high credibility for monetary policy in a situation when Sweden remains outside the monetary union when it starts (Proposition 1997/98:40). According to the Government Bill proposing that Sweden should stay outside the monetary union, Sweden must at least demonstrate its willingness to achieve price stability (and fiscal surpluses) as clearly outside as inside the monetary union (Proposition 1997/98:25).

12 12 decision to stay outside the monetary union, which made it clear that credibility for low inflation policy had to be built at home. 1.3 Challenges for the monetary policy framework The current monetary policy framework has been successful in achieving achieving low inflation, in fact too successful as average CPI inflation was only 1.2 per cent, i.e. 0,8 percentage points below the two per cent target (see Figure 1). This has led to a debate on the Riksbank s policy. It has been claimed that the undershooting of the inflation target has caused unnecessary unemployment, as inflation expectations according to some data have been stable around two percent, which may have led to the re-emergence of a downwardsloping Phillips curve (Svensson 2013). Critics have also argued that the bank has acted asymmetrically in the case of deviations from the inflation target, reacting more to upward than to downward deviations (e.g. Assarsson 2011). An issue raised by these experiences is whether a price level target over a defined period would be superior to the current annual inflation target, since the former target would require the bank to compensate periods of inflation below the target with periods of inflation above it. Another issue is whether the objective of stabilising unemployment around its equilibrium level should be stated explicitly in the Riksbank Act, since this anyway is an implicit objective of the bank. It has also been claimed that a somewhat higher inflation target would be desirable, as this would likely imply higher inflation, and hence a lower real interest rate that would stimulate the economy in a deep recession when the repo rate approaches zero. 11 However, both the political system and the Riksbank have been very reluctant to contemplate such changes. This seems to reflect a worry that such changes could undermine the credibility of low-inflation policies. In the last few years ( ), there has been considerable disagreement in the Riksbank s Excecutive Board. A minority has consistently argued for lower interest rates than the ones decided by the majority. The conflict reflects divergent views on how financial stability should be promoted. In order to contain household debt and house prices, the majority has set the repo rate higher than the level that would be desirable for reaching the inflation target and stabilising unemployment around its equilibrium level. The minority in the Board has 11 These issues have been discussed by, for example, Calmfors (2013a).

13 13 questioned both that financial stability is in danger and that the repo rate is an effective instrument to influence financial stability. 12 The Riksbank Act stipulates that the bank is to promote an effective and secure payments system. This has been interpreted by the Executive Board s majority as a go-ahead for using interest rate policy to promote financial stability. It is a contentious issue whether this interpretation is consistent with the Riksbank Act. The Government Bill on the currently existing Act states that the monetary policy instruments are, however, according to the government s proposal, only to be used to achieve price stability (Proposition 1997/98:40, p. 54). But the Bill also states that in situations where a crisis in the payments system threatens the price stability objective, the monetary policy instruments could be used to prevent a crisis. On the basis of this, the Executive Board s majority has argued that it may be necessary to use the interest rate to prevent excessive borrowing and house price bubbles that could otherwise cause a future financial crisis and a deep recession, which could then make it impossible to attain the inflation target. Even if this view were to be correct, the use of interest rate policy to achieve financial stability is problematic for the accountability of the Riksbank, as almost any (downward) deviation from the inflation target today could be motivated by concerns about future financial crises. The conflict about interest rate policy in the Riksbank s Executive Board reflects the increased awareness of the risks of financial instability in the wake of the on-going international financial crisis and the lack of knowledge of how to best avoid such crises. Financial supervision before the crisis was inadequate almost everywhere. This applies to Sweden, too, where banks were allowed to build up dangerous exposure levels by lending extensively in the Baltic countries. An issue that needs to be solved is how to allocate the responsibility for macroprudential supervision and which additional instruments (such as variable capital adequacy requirements, variable risk weights for various types of loans, loan-to-value regulations and amortisation requirements) to use. 13 There are basically two ways to go. One is to give several agents (the Riksbank, the Financial Supervisory Authority (FSA) and the National Debt Office) joint 12 See, for example, Svensson (2012). 13 It is customary to distinguish between microprudential supervision, that is the supervision of individual banks, and macroprudential supervision, that is the supervision of systemic risks in the financial system, which has to take account also of the interconnections between various financial agents as well as the link between developments in financial markets and the macro economy. The financial crisis starting in 2007/08 revealed that microprudential supervision in Sweden carried out by the Financial Supervisory Authority was not sufficient to monitor systemic risk. This has led to attempts worldwide to develop macroprudential supervision.

14 14 responsibility, but to require them to co-operate in some kind of Macroprudential Supervision Council, as recently proposed by a Government Commission (Finanskriskommittén 2013). The other possibility is to give either the Riksbank or the FSA full responsibility for macroprudential supervision. Joint responsibility of several agents implies serious risks of inefficient policies because no single agent can then be held accountable. Accountability is improved if one agent is given full responsibility. There are arguments both for allocating the responsibility for macroprudential supervision to the Riksbank and for allocating it to the FSA. An argument in favour of the Riksbank is better co-ordination between monetary policy and macroprudential supervision, which both affect macroeconomic developments. Analytical capacity (and tradition) is also much more developed at the Riksbank than at the FSA. Moreover, the Riksbank s independence from the political system could increase the chances that politically impopular decisions on loan-to-value limits or amortisation requirements etc. to prevent excessive credit growth and house price bubbles are indeed taken. On the other hand, there are obvious synergies between the FSA s microprudential supervision of individual financial entities and macroprudential supervision. This would speak in favour of charging the FSA with the responsibility also for the latter task. Alternatively, the Riksbank could get the responsibility for both microprudential and macroprudential supervision. This would, however, entail a huge concentration of power at an institution not directly accountable to the political system (and thus indirectly to voters) in the same way as a government agency like the FSA. One could also argue that macroprudential supervisory decisions are unsuited for delegation to technocrats, as they involve much more of genuinely political distribution concerns (regarding the possibilities of less wealthy households to borrow for acquiring real estate), than short-run interest rate decisions with the aim of minimising variations of inflation around the inflation target and smoothing the business cycle. The recent controversies in the Riksbank s Executive Board has also focused attention on appointment procedures and terms of office. The General Council s selection procedure of new Board members is very non-transparent. This is in contrast to appointment procedures to, for example, the Open Market Committee of the Federal Reserve in the US or the Executive Board of ECB in the euro area, where there are public hearings (in the Senate and in the European Parliament, respectively) with candidates. There is a strong case for such

15 15 parliamentary hearings also in Sweden in order to have a public scrutiny of the qualifications of candidates. The recent disagreements on interest rate policy in the Executive Board have also raised the issue of whether terms of office should continue to be renewable, as is now the case. Renewable terms of office expose Board members to the risk that their monetary policy decisions are influenced by a wish to please the General Council s majority in order to increase the chances of re-appointment The fiscal framework 2.1 The current fiscal framework The current fiscal framework consists of several parts: 1. A top-down approach when the Parliament decides the annual budget. Decisons are taken in two steps. In the first step, the Parliament decides total government expenditure and its allocation among 27 expenditure areas. In the same step, changes in tax rates and various fees are decided. These decisions are taken on the basis of a proposal from the Parliament s Committee on Finance. In a second step, individual expenditure items within each expenditure area are decided. These decisions, which are based on proposals from other committees in the Parliament, cannot change the overall expenditure level in an expenditure area. This decision process ensures that there is an overall decision on total government expenditure so that it (and the budget balance) is not just the outcome of a large number of individual and uncoordinated decisions. Fiscal transparency is promoted by rules on completeness of the budget, which do not allow the use of any extrabudgetary funds, and on gross budgeting, prohibiting the netting out of expenditures against revenues. 2. According to the Budget Act, the government is obliged to propose an annual ceiling for central government expenditure three years ahead. If there is a risk that the ceiling will be broken, the government has to take action or propose actions to the Parliament such that this is avoided. The ceiling puts a limit to budget slippages on the expenditure side of the budget. 14 In 2013, professor Lars E.O. Svensson chose not to run for a second period of office. He had consistently and vehemently argued against the Executive Board majority s interest rate decisions. It is possible that his decision was influenced by an unwillingness of the General Council to re-appoint him (see TT 2013). This might have sent a signal that opposing the majority view in the Executive Board is seen as disqualifying for re-appointments.

16 16 3. The Budget Act also stipulates that the government is to propose a target for general government net lending (a surplus target) to the Parliament. The numerical target is not specified in the Act and can thus be changed by the Parliament without any change in the Act itself. But since its inception in 1997 the target has in effect been one per cent of GDP. 15 The target does not apply to an individual year, but should be met over a business cycle. However, no attempts are made to measure the length of the cycle. Instead adherence to the target is evaluated by a number of indicators: a tenyear backward-looking average of actual net lending, a ten-year backward looking average of cyclically adjusted net lending, a partly forward-looking average of net lending (actual figures three years back and forecasts for the current and three future years), the corresponding partly forward-looking average of cyclically adjusted net lending, and the current structural net lending (which is adjusted for both the cycle and one-off measures). There is no requirement that past violations of the target must be compensated for. On the contrary, the government has made it clear that earlier developments are just a guide to judge the likelihood of meeting the target in the future. 16 This is in line with the idea of tax smoothing, i.e. that it is optimal in order to minimise tax distortions at each point of time to set tax rates so that they can be expected to finance anticipated government expenditure for all future if they are held constant. Then temporary budget shocks are allowed to result in permanent changes in government debt. In line with such considerations, the government has made it clear that, in its view, the budget target does not imply any target for government debt There is a balanced budget requirement for local governments (municipalities and regions). They must budget for an excess of revenues over expenditures. If this requirement is not met, it must be compensated for within three years. Unlike the surplus target for the entire public sector, the balanced budget requirement for local governments does not apply to net lending, which is calculated without accruing investment expenditure, but to the economic result after such accrual. 18 Also in contrast to the surplus target, the balanced budget requirement for local governments 15 Formally, the surplus target was first set at two per cent of GDP. Then savings in the funded part of the government pension system, amounting to approximately one per cent of GDP, were included in general government net lending. This was no longer allowed after a Eurostat decision in 2004, The surplus target was then lowered to one per cent. This means that in effect the target has been unchanged since its introduction. 16 See, for example, Proposition 2009/10:150 and Fiscal Policy Council (2010). 17 Ibid. 18 Such a balanced budget rule is usually referred to as a golden rule of public finances in the international literature; see Fiscal Policy Council 2008).

17 17 applies annually. Possibilities have, however, been introduced for local governments with strong finances to build up balancing accounts (rainy-day funds) in good times that can be activated in cyclical downturns. 5. The state pension system is one with defined contributions, which means that benefits are adjusted to fixed contributions. Pensions are indexed to per capita wage growth, but there is a balancing mechanism that limits the degree of indexation if the long-run financial stability of the system is threatened: this occurs if the capitalised value of contributions plus the assets in existing buffer funds fall below the value of pension liabilities. The balancing mechanism is automatic according to a predetermined formula. Hence, no political decisions are required to ensure the long-run sustainability of the pension system (but new political decisions that endanger sustainability can, of course, be taken) Central government budget decisions are based on a procedure where an annual scope for reforms is calculated by the Ministry of Finance. 20 The scope for reforms is defined as the total sum of permanent tax reductions and government expenditure increases that can be actively decided by the Parliament and that are compatible with the surplus target. This scope for reforms arises because tax revenues grow automatically more or less in line with GDP, whereas government expenditure does not. The reason is that only some expenditures are tied to wages (which grow at about the same rate as nominal GDP), whereas others are linked to wages only after deductions for productivity increases, are indexed to prices (which grow more slowly than wages) or are fixed in nominal terms. Absent discretionary decisions, government net lending thus tends to improve automatically. The Finance Ministry s calculation of the scope for reform forms the basis for the budget discussions within the government. The calculation has usually also been accepted by the main opposition parties A Fiscal Policy Council evaluates the government s fiscal policy. 22 The Council is to assess whether public finances are sustainable in the long run and whether they are consistent with the surplus target and the expenditure ceiling as well as with the 19 See, for example, EEAG (2007) for a brief description of the Swedish pension system. 20 See Fiscal Policy Council (2011). 21 The only exception is when the Social Democrats in 2012 estimated a smaller scope for reforms than the government. 22 See Calmfors (2013b).

18 18 cyclical situation of the economy. The Council consists of six members with either high scientific competence in economics or practical experience of economicpolicy work. An annual report is published in May each year, about one month after the government has presented its Spring Fiscal Policy Bill. The report is formally addressed to the government, but is also the subject of a public hearing in the Parliament s Committee on Finance. In addition, the public finances are regularly monitored also by the National Institute for Economic Research, the Office for Budget Management and the National Audit Office. 2.2 The establishment of the fiscal framework The major part of the fiscal framework was established as a direct response to the economic crisis of the early 1990s, when very large fiscal deficits emerged (Figure 2). 23 In 1993, the deficit was 12 per cent of GDP. Government debt was increasing fast (Figure 3) and government bond yields were high. The then liberal-conservative government made it one of its top political priorities to stop the accumulation of debt in This objective received an even higher priority when the Social Democrats came to power in 1994 and formulated a fiscal consolidation programme. It contained clearly stated objectives: first (November 1994) that government debt should be stabilised as a share of GDP latest in 1998 and later (April 1995) that a balanced budget (zero general government net lending) should be achieved in 1998, and that government debt as a share of GDP should be stabilised already in In June 1995, the government set the objective that government net lending were not to exceed three per cent of GDP (a convergence criterion for joining the monetary union) in In its first Budget Bill in 1995 (Proposition 1994:95:100) the Social Democratic government also announced that the fiscal consolidation programme was to be complemented by institutional changes of the budget process. The first changes to be imposed were the central government expenditure ceiling and the top-down budget process, which were used for the first time in The consolidation programme was very successful and the set objectives were achieved (see Figures 2 and 3). The success was to a considerable degree attributed to the formulation of clear and well-publicised objectives. This served as an important inspiration for the formulation of the surplus target, which was decided in The numerical level chosen - 23 See, for example, Finans- och penningpolitiskt bokslut för 1990-talet (2001) or Fiscal Policy Council (2008).

19 19 one per cent of GDP - was based on the insight that the strains on future public finances arising from an ageing population could be eased through pre-funding (saving in advance) and that fiscal surpluses would increase the scope for countercyclical fiscal policy in downturns. 24 In 2000, the described reforms of the fiscal framework were complemented by the balanced budget requirement on local governments, which was motivated by a desire to avoid that fiscal discipline in the public sector would be jeopardised by a lack of discipline in local governments. As with the reforms of the monetary framework, academic input was important for the reforms of the budget process, but the impact was more indirect. The direct influence came from civil servants in the Ministry of Finance, based on their earlier experiences of budget work. A key role was played by a study, made by a civil servant, which found that the Swedish budget process was very weak in an international comparison and recommended a stricter framework including a top-down budget approach and a government expenditure ceiling (Molander 1992). 25 A main conclusion in the study was that earlier economic crises in Sweden depended to large extent on an inability to contain expenditures in good times (Molander and Holmquist 2013). The study was inspired by an earlier academic analysis of budget processes in EU countries, which found that they had an important effect on fiscal outcomes (von Hagen 1992). The conclusions in the Molander study were endorsed by the Lindbeck Commission (1993), which was important for gaining acceptance for these ideas in the political system. They were first adopted by the Parliament in its work on the budget and subsequently by the government in its internal preparation of the budget. 26 European influences were also very important for the introduction of the stricter fiscal framework. The fiscal crisis and the consolidation process in the 1990s coincided in time with the formulation of the EU fiscal rules in the Maastricht Treaty and the stability pact. 27 Paradoxically, it appears that Sweden took these rules much more seriously than the member states joining the monetary union. The government stated that Sweden had to demonstrate its ability to establish a fiscal surplus at least as clearly outside as inside the monetary union 24 Ibid. See also Stabilisation Policy in the Monetary Union (2002). 25 The study was commissioned by the Expert Group for Public Economics (ESO), an independent committee attached to the Ministry of Finance, with the remit to make an independent contribution to expanding and deepening the knowledge data available to future socio-economic and fiscal policy decisions. See also Molander and Holmquist (2013) for a recent discussion of how the stricter fiscal framework was adopted. 26 The main ideas were set out in Talmanskonferensen (1994) and Molander et al. (1995). 27 The Maastricht Treaty was agreed in 1992 and adopted in The discussions on the stability pact started in 1995 and the pact was finalised in 1997.

20 20 (Proposition 1997/98:25). To establish a sufficient margin to the EU three-per-cent-of-gdp deficit ceiling in normal times, in order to avoid violations in downturns, was also an explicit deliberation behind the formulation of the surplus target. 28 The first major fiscal reform to be decided in the 1990s was a five-party agreement on a new pension system with defined contributions (see above) in 1994 between the then four liberalconservative government parties and the Social Democrats. The background was a growing awareness over time that the earlier pension system was not sustainable in the long run, as there was not a sufficient link between the benefits and the incomes that should finance them. The earlier system was one of defined benefits and therefore it could be considered a violation of an existing implicit contract between generations if benefits had to be tampered with, as they were several times in the 1980s and again in 1992 as part of measures to improve the fiscal balance. It was also realised that the earlier system provided weak incentives for a long working life, as benefits were based not on lifetime earnings, but only on earnings during an individual s best 15 years. Pensions were traditionally a politically highly polarised area. The earlier pension system was introduced by a Social Democratic government in 1959 in conflict with the liberalconservative opposition. Although the sustainability problems of this system were highlighted by a Government Commission working between 1984 and 1990, the political parties had not been able to agree on pension reform (Pensionsberedningen 1990). It was the acute economic crisis in the early 1990s, and the fears that a period of high unemployment and low growth would further undermine the pension system s sustainability, that made the five-party agreement possible (Lundberg 2001). It can be seen as a continuation of earlier agreements in 1992 between the liberal-conservative government and the Social Democrats on how to cope with the acute economic crisis. After the fiscal reforms in the 1990s, no further changes were implemented for some time. There were fiscal surpluses and the government debt was gradually reduced (see Figures 2 and 3). However, the liberal-conservative government coming to power in 2006 implemented additional reforms. They were not motivated by any acute fiscal problems. Instead, they were responses either to academically based proposals or to a general political desire (or at least to the Finance Minister s desire) to further improve the fiscal framework. The latter motive 28 See, for example, Stabilisation Policy in the Monetary Union (2002), Utvärdering av överskottsmålet (2010) and Proposition 2009/10:150.

21 21 should be seen in the light of a strong political will in the liberal-conservative government to show that it could handle the public finances well (the earlier liberal-conservative coalition governments in and presided over large deteriorations of the fiscal balance). 29 The first reform was the establishment of the Fiscal Policy Council in In recent years, there has been a strong international trend, particularly in the EU, to set up such national fiscal watchdogs. Here, Sweden was quite early and the Swedish Council has to some extent served as a role model for similar institutions elsewhere. The idea of such independent fiscal monitoring institutions first appeared in the international academic debate in the 1990s as an attempt to find ways of transferring the benefits of independent policy-making in the area of monetary policy to that of fiscal policy. In Sweden, the idea was picked up by a Government Commission with the remit of analysing fiscal policy in the event of membership in the monetary union (Stabilisation Policy in the Monetary Union 2002). Based on a background report by Wyplosz (2002), the Commission proposed the establishment of an independent fiscal watchdog. 30 The proposal was rejected by the Social Democratic government at the time, but was more popular with the liberal-conservative opposition. It was endorsed by the then chief economist of the Moderates (the Swedish Tory Party and the largest one of the liberal-conservative parties), Anders Borg (Borg 2003). In 2006, he became Minister for Finance in the liberalconservative government and then pushed through the establishment of the Council. This was done although the parties of the left (Social Democrats, the Greens and the Left Party) opposed it on the grounds that unelected experts would get too much power and that the Council s analyses were bound to have a liberal-conservative bias. However, these parties changed their views gradually and in 2011 entered an agreement with the government regarding the Council, extending its remit also to analysis of income distribution issues. The main reasons for the change in views seems to have been that the Council quickly earned a reputation for non-partisanship and competence as well as a realisation that critical 29 It is also well-known from empirical research in political economy that budget deficits tend to be larger with coalition governments than with one-party governments (see Persson and Tabellini 2002). 30 See, for example, Calmfors (2003) and Debrun et al. (2009) for surveys of proposals on independent fiscal monitoring institutions. Existing such institutions have been surveyed by the European Commission (2006), Hagemann (2010) and Calmfors and Wren-Lewis (2011). See also OECD (2013) for a list of OECD notes on such institutions in various countries.

22 22 assessments of government policies and suggestions for improvements are in the interest of the opposition. 31 Other recent changes in the fiscal framework have been made on the government s own initiative, largely without any major input from more academic thinking. They seem to have reflected a genuine interest by the Finance Minister (Anders Borg) in fiscal framework issues, an interest which was further stimulated by the soverign debt crises in the euro area. The Spring Fiscal Policy Bill in 2008 set out clear principles for calculating the annual scope for reforms (see Section 2.1), based on the surplus target and the cyclical situation, as a means of imposing more discipline on the budget process (Proposition 2007/08:150). Before 2011, it was not obligatory for the government to propose an expenditure ceiling (although it had always been done after the possibility was introduced in the Budget Act of 1996) and the surplus target had no formal legal backing. From 2011, the Budget Act makes it obligatory for the government to propose both an expenditure ceiling and a surplus target to the Parliament. To sum up, the establishment of a stricter fiscal framework took place in two steps. The most important reforms were done in the second half of the 1990s and were a direct response to the then fiscal crisis. The proposals on reforms came mainly from experts inside the Ministry of Finance. The formulation of fiscal rules at the EU level provided an important source of inspiration, not least because there was a wide consensus that Sweden should live up to these rules and hence be able to decide itself from a position of strength whether or not to join the monetary union. There was a second round of reforms, though less pervasive, from It was not triggered by any acute fiscal problems in Sweden, but seems instead primarily to have been motivated by arguments of principle on how to strengthen the framework further. In this phase, academic reasoning played a more direct role for the establishment of a fiscal watchdog. 2.3 Challenges for the fiscal framework On the whole, the fiscal framework has worked well and delivered fiscal discipline. Figure 2 shows that the surplus target of one per cent of GDP has for most of the time been exceeded and Figure 3 that government debt has fallen strongly. It is noteworthy that this has occured without any formal sanction procedures being in place: instead the respect for the fiscal framework seems to have been based on a political consensus never again to get into a fiscal 31 Positive assessments of the Council s activities were made, for example, by the IMF (2010) and the OECD (2011). See also Calmfors (2013b). The author of this chapter was the Council's first chair in

23 23 crisis situation requiring harsh consolidation measures as in the 1990s (EEAG 2011; Calmfors 2012a,b). There is no pressing need to reform the framework it in order to deliver better outcomes. Still there are several issues to discuss. The multitude of indicators for evaluating whether or not the surplus target is achieved (see Section 2.1) introduces an element of ambiguity which could potentially be misused by the government. Therefore, the Fiscal Policy Council (2010) has proposed a clearer definition of the target. Another potential problem is that it is unclear how (properly defined) violations of the target are to be dealt with (a clear contrast to the automatic correction mechanisms that the eurozone countries are now introducing in their national legislation). The Fiscal Policy Council (2013) concluded that such a violation is now (2013) occurring (associated with the long period of low resource utilisation in the wake of the international economic crisis) and wanted a transparent explanation on how the government plans to handle the situation. Such a procedure could be institutionalised, for example, by requiring the government to send a formal letter to the Parliament in such situations. Another issue concerns the numerical value of the surplus target. If government net lending is one per cent of GDP on average, government net financial wealth will, under plausible assumptions, continue to increase from the current level of around 20 per cent of GDP and ultimately converge to a level somewhere between 30 and 50 per cent (Fiscal Policy Council 2012; SNS 2012). 32 It is not obvious that such a large precautionary buffer is needed in the event of future fiscal crises. There is currently an ongoing debate on whether the surplus target is too ambitious. The deep economic downturn in 2009 raised the issue of whether the government expenditure ceiling should have an escape clause, allowing the automatic stabilisers on the expenditure side to work fully in such a situation, as was proposed by the Fiscal Policy Council (2009). This would make it unnecessary for the government to resort to creative accounting to prevent violations of the ceiling (as happened in 2009/10). But the government has been very opposed to this on the grounds that it could weaken the ceilings s credibility, preferring instead to deal with contingencies in a discretionary manner. This would, however, in the view of the author of this chapter be much more problematic, as no ex ante limits for violations are then set. 32 The exact level will depend on how the value of the stock held by the government sector develops.